Sie sind auf Seite 1von 29

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality

Elizabeth Barnes

University of Leeds e.j.barnes@leeds.ac.uk

In this paper, I argue for a new way of characterizing ontological emergence. I appeal to recent discussions in meta-ontology regarding fundamentality and de- pendence, and show how emergence can be simply and straightforwardly charac- terized using these notions. I then argue that many of the standard problems for emergence do not apply to this account: given a clearly specified meta-ontological background, emergence becomes much easier to explicate. If my arguments are successful, they show both a helpful way of thinking about emergence and the potential utility of discussions in meta-ontology when applied to first-order metaphysics.

In this paper, I argue for a new way of characterizing ontological emergence. 1 I appeal to recent discussions in meta-ontology of funda- mentality and dependence and show how emergence can be simply and straightforwardly characterized using these notions. 2 I then argue that many of the standard problems for emergence do not apply to this account: given a clearly specified meta-ontological background, emergence becomes much easier to explicate. If my arguments are successful, they show both a helpful way of thinking about emergence and the potential utility of discussions in meta-ontology when applied to first-order metaphysics. The specific commitments of emergence are opaque. Emergentists maintain that the parts of a system, through their collective activity, can sometimes give rise to an entity which is quite distinct — in terms of its structure, its causal powers, its ontological makeup, etc. — from the parts of that system, or from anything those parts compose.

1 There is a separate notion, often employed in philosophy of science, of epistemological emergence which does not concern me here. I am interested in emergence as a metaphysical thesis: the prospect of emergent entities (with ‘entity ’ being understood as either an object or a property).

2 I do not by any means wish to assert that appealing to fundamentality and/or dependence is novel to discussions of emergence. Rather, the way in which I make this appeal — tying the explication of emergence specifically to recent debates in meta-ontology — is to my knowledge new, and presents an as-yet-unexplored way of thinking about emergence.

Mind, Vol. 121 . 484 . October 2012

doi:10.1093/mind/fzt001

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

874 Elizabeth Barnes

The resultant entity, in these special cases, is emergent. Things get tricky, though, when we try to understand exactly what is meant by this notion of emergence. It is often couched in metaphors — for ex- ample, an emergent entity is ‘something new’ (as in, e.g., Marras 2006 ), something which is ‘over and above’ its parts (as in, e.g., Crane 2001) — and it is not clear how to cash those metaphors out. One thing that does seem clear, however, is this: to claim that some entities are emergent is to make a claim about the world’s structure.

The emergentist is saying that not everything about the world’s struc- ture can be explained in terms of smaller parts building into larger, more complex wholes. Sometimes the collective activity of the smaller parts produces not just a complex thing made up of those smaller things, but also something fully distinct from (‘emergent’ from) those parts, their sum, and anything they compose. And that is a claim about how the world is structured.

If such claims are going to be made clearly, they need to be made in

the context of an explicit and coherent ontological (or perhaps more accurately, meta-ontological) framework. My claim is that much of the mystery about emergence arises from having no clear statement of what its meta-ontological background should be. In this paper, I will show how a specific meta-ontological framework can be used as a background for understanding emergence. Within this framework, I will argue, many of the standard puzzles and problems for emer- gence — and in particular, the worry that emergence is ‘mysteri- ous’ — can be resolved. Emergence, whether or not you think there is any reason for endorsing it, is not on this interpretation mysterious.

Appeal to the concepts familiar from the literature on meta- ontology, I argue, allows us to make the discussion of emergence

more rigorous (I outline one way of doing this — there are many others). It allows us to provide a full characterization of a specific option in conceptual space, one which looks like it could plausibly be described as ‘emergence’. We can then present this view to the defender of emergence and ask: ‘Is this your view? If not, why not?’ The hope is that the rigour provided by a meta-ontological framework will enable some traction on the all-too-slippery idea of emergence, and thus provide for some progress in the debate.

I begin by sketching a structural distinction — which I will call a

‘fundamentalist ontology ’ — between the fundamental and the deriva- tive (Sect. 1 ). I will then argue that this distinction can be pulled apart from a further distinction — that of the ontologically dependent and the ontologically independent (Sect. 2). Separating these two

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 875

distinctions allows for the characterization of emergent entities as those which are fundamental but not independent (Sect. 3 ) and co- heres nicely with standard (as well as some non-standard) examples of emergence (Sect. 4). I will argue that this way of explicating emergence avoids many of the traditional problems associated with emergence, which often presuppose a connection between fundamentality and a hierarchy of ontological levels — a connection that the fundamentalist ontology explicitly rejects (Sect. 5 ). I will not argue for the ontological framework — the ‘fundamental- ist’ ontology — that I explain in subsequent sections. Nor will I argue that there is emergence. If you are committed to emergence, you can read what follows as giving (defeasible) motivation for a fundamen- talist ontology (since a fundamentalist ontology helps explicate emer- gence). Likewise, if you are committed to a fundamentalist ontology, and think that it is at least an open (epistemic) possibility that there is emergence, you could count as one more advantage for ontological fundamentalism that it can help explain emergence. But this paper will not give you independent arguments either for a fundamentalist ontology or for emergence. Rather, my aim is to show that establishing a specific ontological background like ontological fundamentalism can sometimes help to resolve longstanding problems — in this case, some vexed questions about the metaphysics of emergence.

1. The fundamentalist ontology

The basics of an approach which I call the fundamentalist ontology has recently been defended by, inter alia , Kit Fine ( 2001), John Heil ( 2003), J. R. G. Williams (2010 ), Jonathan Schaffer (2010 and 2010b), 3 and Ross Cameron (2008 and 2010 ). 4 The fundamentalist ontology maintains a single crucial ontological distinction: the divide between what exists derivatively and what exists fundamentally.

3 Or, better, this approach is cited as one of the possible interpretations of monism which Schaffer proposes, though remains neutral on. Schaffer argues that only the world is absolutely fundamental, and he says that one way of interpreting this is that everything else is simply derivative, and thus on a par. He remains neutral, however, on whether this is the correct interpretation (as opposed to saying that there are degrees of fundamentality, with the world as the only absolutely fundamental thing, but with other entities being more or less fundamental than one another).

4 There are different ways of explaining the distinction — e.g. Fine 2001 characterizes it as the distinction between what exists and what exists in reality. I will proceed with the funda- mental/derivative distinction here, but translation to other locutions should not be problematic.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

876 Elizabeth Barnes

Fundamentality, as deployed in metaphysics, is often considered a matter of degree: for example, molecules are less fundamental than electrons, but more fundamental than tables and chairs. 5 But, in contrast to a picture wherein certain things can be more or less fundamental than other things (even if they are not absolutely funda- mental), on the fundamentalist view things are either fundamental or they are not, in which case they are derivative. Fundamentality does not come in degrees. That is, everything that is not fundamental exists in the same manner; rather than having degrees of fundamentality (which will vary inversely with the degree of complexity), everything that is not fundamental simply has its existence derivatively from what exists fundamentally. To get clearer on this, more needs to be said about the notion of fundamentality invoked here. I will take fundamentality to be a meta- physical primitive, and as such I will not attempt to give a definition of it. I can, however, gloss it in a number of ways that may be helpful. Take, for example, the familiar theological metaphor: the funda- mental entities are all and only those entities which God needs to create in order to make the world how it is. So if God wants to create a world w , the fundamental entities will be the entities necessary and sufficient for God to create in order for her creation to count as a creation of w . Likewise, if she changes her mind and decides to create w * instead, she will alter her creation by changing what fundamental entities she creates; she need change her creation only in fundamentals in order to make it a creation of w * rather than w . So, for example, if God decides that she wants a world with a single complex object composed of two mereological simples, she would simply have to create the two mereological simples. 6 She would not have to engage in a further act of creation: ‘let there be a complex object’. By setting the composition relation and what simples exist, she gets the (derivative) complex object for free. 7 Alternatively, fundamentality can be cashed out in terms of truth- makers. Entities which are fundamental are those which truthmake their own existence, and which are capable of serving as truthmakers

5 See especially Heil 2003 and Schaffer 2003 for critical discussion.

6 And arrange them appropriately, if restricted composition is true.

7 Something like this will not work as an analysis of fundamentality, however, because we should not rule out by definition that there are necessary fundamental existents, or necessary connections between fundamental existents.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 877

for the existence of other (derivative) entities. 8 So, for example, sup- pose that a trope bundle theory is true, and that tropes are funda- mental but tables are not. The tropes truthmake their own existence; for each individual trope, you need appeal to no ontology other than that trope itself for the truth of ‘that trope exists’. Contrast this to the existence of a table. The table exists, but the truthmaker for its exis- tence is not the table itself, but the tropes it is composed of (plus relations between them — be it mereology or compresence). Each of the tropes truthmakes its own existence, and they also jointly truth- make the existence of the table. 9 Derivative entities will then be characterized in simple contrast to fundamental ones. An entity is derivative just in case it is not funda- mental. Derivative entities are those which, as it were, derive their existence from the fundamental entities. Couched in the previous metaphor, derivative existents are those which God does not need to include in her ‘ontological shopping basket’. She can create them simply by creating the fundamental existents which they are derived from. Likewise, the derivative entities are those which are not the truthmakers for their own existence. For any derivative entity x , ‘x exists’ is true but made true, not by x, but by some collection of fundamental entities y 1 y n . Being derivative, in this characterization, is irreflexive, asymmetric, and transitive (but only trivially so, since you never get chains of derivativeness). Derivative entities are derivative only on fundamental entities, never on other derivative entities. 10 That is, on the alternative picture in which fundamentality comes in degrees we have a ‘hierarchical’ structure, as illustrated in Fig. 1 (where things which are Fundamental are absolutely fundamental, things which are Derivative 1 are less fundamental than the Fundamental things but more fundamental than the Derivative 2 things, and so on); whereas on the fundamentalist approach we have the situation as shown in Fig. 2 .

8 See Heil 2003 and Cameron 2008 for discussion of this approach.

9 If you find both of these glosses unhelpful, you can think about fundamentality in terms of explanation. So, for example, the derivative entities are explained by the fundamental entities, or exist in virtue of the fundamental entities, whereas the fundamental entities do not exist in virtue of anything. I worry, though, that explanation-talk is too contextually shifty to serve as a characterization of fundamentality.

10 Though there are various ways of cashing out the notion of ‘derivative on’. See n. 19 .

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

878 Elizabeth Barnes

Fundamental Derivative 1 Derivative 1 Derivative 2 Derivative 2 Derivative 2 Derivative 3 Figure 1
Fundamental
Derivative 1
Derivative 1
Derivative 2
Derivative 2
Derivative 2
Derivative 3
Figure 1 : The hierarchical view of fundamentality
Fundamental
Derivative
Derivative
Derivative
Derivative

Figure 2 : The fundamentalist view of fundamentality

There are different ways to characterize the existence of the deriva- tive entities — the picture can be either deflationary or inflationary. The former interpretation would say something to the effect of ‘fun- damental things exist and derivative things exist, but only fundamen- tal things really exist’. 11 Or ‘of course we can say true things about the derivative existents, and in that loose sense they exist, but the only

11 See Cameron 2008 and Cameron 2010 for explanation of this more deflationary approach. Cameron argues that the truth in natural language of ‘ x exists’ does not require that we include x in our ontology. Cameron adopts the Lewis-Sider account of meaning and quantification, claiming that meaning is determined by both use and naturalness. But use can sometimes trump naturalness, if common usage is very unnatural. So it might be that the quantifier expressed by ‘exists’ in English is very unnatural. If that is the case, there can be true sentences of English (and other natural languages) of the form ‘x exists’ which do not ‘carve the world at its joints’ — that is, they do not correspond to the non-linguistic structure of reality. But this does not preclude there being some other language (though not a language anyone speaks — call it ‘Ontologese’) which is perfectly natural and whose quantifiers do carve the world at its joints. It is the true sentences of the form ‘ x exists’ in this language, Cameron argues, that the metaphysician should be interested in. And a complete inventory of such existence facts will be sufficient for the truth of the true sentences of English (even though English quantifies over things Ontologese does not, and possibly vice versa).

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 879

things that make up the world in some metaphysically privileged sense are the fundamental things.’ The latter interpretation, in contrast, would claim something along the following lines: ‘of course both fundamental and derivative entities exist, and of course they really exist (whatever that means), there is just a difference in how they exist or what the nature of their existence is’. 12 For the purposes here I will remain neutral on this question, as nothing in the subse- quent arguments will hang on whether an inflationary or deflationary approach to derivative existence is adopted.

2. Ontological dependence

I will appeal to one other basic distinction: that between entities which are ontologically dependent and those which are ontologically inde- pendent. I will argue that the two distinctions are distinct and come apart. This potential separation between fundamentality and indepen- dence will be what gives rise to the basic characterization of emergence that follows. The basic idea behind dependence is often spelled out (unhelpfully) as something like this: an entity x is dependent on entities y 1 y n just in case x is sustained by the y s 13 (or perhaps both caused and sustained by the y s — but the account of dependence I will give is non-causal). The difference between something simply being caused by something else and something being sustained by something else will do a lot of work in the understanding of dependence in play here, so it is best to spell it out in more detail. I suspect that, like fundamentality, depen- dence is primitive, so I do not intend what follows as an analysis. Think of it, rather, as a helpful explanation that generalizes from basic cases: what I am concerned with is giving an intuitive gloss on the idea I am labelling ‘dependence’. There is a sense in which a person is dependent on her parents for her existence: her parents cause her to exist, and had her parents not existed she would not have existed. But this is a very weak sense of dependence: it does not demand that anything exists now whose exis- tence you depend on — it only requires that at some time or other,

12 This is the interpretation favoured by, inter alia , Jonathan Schaffer. See his 2010 and

2010 b.

13 Or ‘the existence of x is sustained by the existence of the y s’, if you think that is an important difference.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

880 Elizabeth Barnes

something(s), y 1 y n , existed, such that (roughly) had they not existed then you would not exist now. There is, though, a stronger sense of dependence — that had between, say, a complex object and its parts — which is, I take it, what is being suggested by talk of something being ‘sustained’ by some other thing(s). The contrast is that, rather than merely being counterfactually dependent on the existence of something in its past, the object is dependent at each moment of its existence on the exis- tence of something which exists at that very time. Take the case of a complex object dependent on its parts. It is not enough for the object to exist at t that at some previous time t n there were some simples arranged object-wise. Rather there must be simples arranged object- wise at t in order for the object to exist at t. The stronger notion of dependence, which I will call ontological dependence 14 can be under- stood as:

(OD) An entity x is dependent iff for all possible worlds w and times t at which a duplicate of x exists, that duplicate is accompanied by other concrete, contingent objects in w at t

That is, in any situation in which there is something exactly like x , you’ve got to have other things existing alongside it. 15 You cannot ever just have x by itself. And in that sense, x depends (at every moment of its existence) on other things. So, for example, even if essentiality of origin is true, you are not, in this stronger sense, dependent on your parents. Your existence requires the existence of your parents, but it is not the case that for every time at which you exist, your parents must exist at that time. And there could be an exact copy of you in a world that does not contain exact copies of your parents (that is why (OD) is formulated in terms of duplicates). Contrast this to the relationship you bear to

14 From here on, unless specifically noted, ‘dependence’ will be understood in this stronger

sense.

15 So any cases where the requirement of (OD) is met apart from those cases where a thing depends on its parts will be cases where Humean recombination fails. (OD) will likely be unattractive to those with strongly non-Humean views of modality, though it is not implau- sible to think that if there are lots of necessary connections in the world there are a lot more dependencies than we might originally have suspected. In any case, I am not optimistic about getting a definition for dependence — especially a modal definition — that makes all varieties of modalist happy, so I am untroubled if (OD) needs to presuppose some kind of Humeanism.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 881

the parts which compose you: at every time you exist, some collection of parts must be there to make up your body. And anything which is an exact copy of you had better include copies of those parts. You are dependent on the parts of your body, but not on your parents. In contrast, a duplicate of something like a mereological atom could exist unaccompanied by anything. That is, you could have an exact copy of the atom without needing to have a copy of anything else to support its existence. Notice also that what is being explained by (OD) is not the depen- dence of x on the y s. Rather, it is the dependence of x simpliciter . 16 A complex object is dependent because you cannot have a duplicate of it without having some things which are its parts. But it would be at best misleading to say that it is dependent on its particular parts. Nothing about dependence encodes the essentiality of constitution — the com- plex object requires the existence of some parts or other, but not the particular parts which in fact compose it. (OD) is restricted to contingent entities because it should not come out that, if there are necessary existents (e.g. abstracta), everything trivially depends on those necessary existents. 17 If (OD) was aiming for an analysis of dependence, you might hope that this would fall out of the definition, rather than be packed into it. But since I am here only trying to provide a useful gloss on the idea, I do not consider this too much of a problem. The basic distinction is just this: if God took away everything dis- tinct from, for example, a table, she would by that very act have to take away the table. She has taken away the simples that compose the table, and so the table goes with them. The table is ontologically dependent. In contrast, if she took away everything distinct from one of the

16 For convenience, I will sometimes use the locution ‘dependent on’, but nothing signifi- cant should be read into that. The key notion is simply dependence simpliciter . I take ‘depen- dent on’ to mean something roughly like the following: x is dependent on the y s iff x is dependent because it is part of its intrinsic nature that it bears relation R to things intrinsically like the y s.

17 This allows that abstracta can depend on concreta, but if we also assume that abstracta exist necessarily, it rules out that anything (including abstracta themselves) depends on abstracta. I do not find this problematic. But if you think that, for example, necessarily existing abstracta can depend on other necessarily existing abstracta, no sort of modal definition is going to work. Likewise if you think there could be necessarily existing concreta, and that some of these necessary concreta are dependent on other necessary concreta: again, modal definitions will not work if you want to capture something like this. On my account, it falls out that if something is necessary, then it is not dependent — and I am happy with this.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

882 Elizabeth Barnes

simples, nothing in that very action commits her to taking the simple as well. One of the simples can get by without the others (and without the table) in a way the table cannot get by without the simples. The simples are ontologically independent. Something is ontologically independent, on this reading, just in case it is not ontologically dependent. If the existence of x does not, at each moment of its existence, rely on some other entity or entities, then x is ontologically independent. So the ontologically independent entities are those capable of ‘lonely existence’, in the Lewis-Langton (1998) sense.

3. Fundamental dependents

So we now have two basic ontological distinctions: the divide between the fundamental and the derivative, and the divide between the depen- dent and the independent. It is tempting to think that these two distinctions are more or less glosses on the same basic idea, but I want to dispute this. I will argue that the doctrine of emergence 18 can be captured in the simple thought that the two distinctions cut across each other. My central thesis is this: that there is ontological emergence is the claim that some things which are fundamental are not ontologically independent. If fundamentality and independence are separate distinctions which cut across one another, then there is space for four possibilities: the fundamental independent, the fundamental dependent, the derivative independent, and the derivative dependent. That there could be fun- damental independent entities is hardly controversial. If there were mereological atoms, for example, they would plausibly be both funda- mental and independent in the sense given above. Derivative depen- dent entities are not a hard sell either: plausibly macroscopic objects like tables fall into this category. It is the other two combinations that are problematic, but I think there is conceptual space for both. I am going to argue, in subsequent sections, that fundamental dependent entities plausibly fit the characterization of emergent entities; but before I get to that I will say something about why there might be derivative independent entities. Consider, for example, numbers. On a Platonic metaphysics of mathematical entities, numbers are fundamental; but on others, they are derivative — and while on some metaphysics that take them to be derivative they might be held to be dependent (on, for example,

18 Or, at least, one version of it. A lot of things get labelled ‘emergence’.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 883

the structure of infinitely many concrete entities), there are plausible views whereby numbers are derivative but independent. Consider mathematical trivialism, as defended recently by Agustı´n Rayo (2009 ). Rayo suggests that mathematical truths are trivial in the sense that their truth imposes no demand on the world. So ‘there is

a prime number between 4 and 6 ’ is true, but nothing is required of

the world in order for it to be true (as opposed to ‘there is an ele- phant’, whose truth requires there to be an elephant). While Rayo does

not speak in these terms, it is natural to take this as a view according to which numbers are derivative: they are no part of fundamental reality. 19 Since the truth of sentences proclaiming the existence of numbers makes no demands on the world, numbers need be no part of God’s ontological shopping basket if she wants to make a world where the truths of maths are true. In Fine’s terminology, numbers exist on this view, but they do not exist ‘in reality ’; in the language of the truthmaker theorist, numbers need not be included amongst one’s ontology to do any truthmaking work, because the truths of maths are trivially made true. So this is a metaphysic according to which numbers are derivative; but are they dependent? Plausibly not. If we assume that a duplicate of

a number is just itself, it seems that numbers could exist unaccompa-

nied by any contingent things (unless a world empty of concreta is impossible), so they meet the requirement of (OD) for independence. Of course, (OD) was not meant as an analysis of dependence, but merely a litmus test for dependence in paradigm cases, and one might think necessary existents are precisely not paradigm cases; that is, if the definition was going to break down, it would be when dealing with necessarily existing dependent entities. That may be so. Nevertheless, on this view it does not seem correct to say that numbers are dependent. Why would they be? This is not a view on which talk about numbers is talk about some structure(s), either abstract or concrete, possible or actual; it is not a view on which numbers are properties or sets and can therefore plausibly be said to be dependent

19 The obvious question, then, is: What are they derivative on? Nothing, I claim. Some things are derivative on others, some are merely derivative. We characterized fundamentality as follows: x is fundamental iff x makes true [ x exists]. So for y to be derivative is just for [ y exists] to be true, but y not to make it true. But there are two ways that can be the case. It can be the case because something other than y, z say, makes it true, in which case y is derivative on z, or it can be the case because nothing makes it true that [ y exists], in which case y is simply derivative, without being derivative on anything. The latter, it seems, is the position of the trivialist. See, for discussion, Cameron 2010 b.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

884 Elizabeth Barnes

on their instances/members. This is a view according to which num- bers are an ontological free lunch: their existence makes no demand on the world, a fortiori no demand that there be some entities on which they depend (in the way that the existence of a table makes the demand on the world that there be parts which compose it). Numbers, on this view, look independent. But they are none the less derivative, because the fact that they exist does not demand that they be included in the fundamental ontology (since it does not demand anything). And so if this view is correct, numbers are derivative independent entities. Now, I am not arguing here that it is correct. But it seems to me to be coherent, and so we should at least admit the conceptual possibility of independent derivative entities. And once we are happy with the fundamental and the independent coming apart in that direc- tion, we should drop any lingering suspicion against the coherence of it coming apart in the other direction — that is, with the coherence of the idea that there are fundamental dependent entities. And the class of fundamental dependent entities, I will argue, is the class of emer- gent entities. And so we can make divisions as outlined in Table 1 . 20 As I will explain below, I think that all entities plausibly described as emergent should be characterized as both fundamental and depen- dent, and I think that an entity cannot be both fundamental and dependent without being considered emergent. Thus I think that the following holds:

An entity x is ontologically emergent iff x is fundamental and dependent

I will first explain, by some general remarks and a few examples, why I think (OE) captures the basic idea of emergence. I will then argue that the characterization of emergence given by (OE) avoids many of the standard problems faced when emergence is couched in the more familiar framework of a hierarchical levels ontology. The central idea of ontological emergence has generally been glossed as getting ‘something new’ which is ‘over and above’ its parts. An emergent entity is thus taken to be a substantial ontological

(OE)

20 I am not committed to the particular examples given in Table 1 . One might, for example, argue that there are fundamental complex objects, or derivative simples, etc. The examples are just meant to illustrate the framework — and what is important for present purposes is that each of the four categories is potentially occupied by some things, and that emergent entities are whatever goes in the top right box.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

Table 1 :

Emergence and Fundamentality 885

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

 

Independent

Dependent

Fundamental

Mereological simples Numbers (and other necessarily existing abstracta?)

Emergent entities

Derivative

Complex objects,

artefacts, etc.

commitment — you do not get emergent things for free. Yet emergent entities are likewise characterized as not being among the very basic building blocks of the world (whatever they may be). Here, then, is where the two distinctions can do some work. An emergent thing is a robust ontological commitment — ‘something new’, distinct from the sum of its parts. And so, ostensibly, it is one of the things God would have to create in order to make the world how it is. If the table is simply a complex object made up of simpler parts, God can create the parts, arrange them table-wise, and get the table for free. But if the table is emergent — something distinct from the sum or arrangement of its parts — then God cannot just create the parts and arrange them table-wise. She has to create the table as well. Likewise for the truthmaker gloss on fundamentality. Whatever reasons pushed you towards emergence would also likely push you towards the idea that you need emergent entities in your ontology as truthmakers. If the emergent entity is really something new, some- thing over and above its components, then those components will not be sufficient truthmakers for its existence. You will need the emergent thing itself (and it looks plausible to think that the emergent entity will serve as the truthmaker for its own existence). Thus emergent entities look plausibly characterized as fundamental. They are part of the real stuff, the core stuff, in the world. They are not the stuff you get for free. If a basic distinction between fundamental and derivative is in place, emergent entities are best understood as fundamental. But emergent things are clearly not ontologically independent. They depend on the entities from which they emerge. That is, though the emergent entity is ‘something new’, the existence of the ‘new’ entity must be both caused and sustained by the collective activity of other entities. Otherwise, the entity in question would not be properly char- acterized as ‘emerging’ from anything. What about the other direction of the biconditional? Why charac- terize all cases of fundamental dependent entities as emergent? I am

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

886 Elizabeth Barnes

less committed to this claim, but I do find it plausible. Emergent entities have typically been characterized as straddling some sort of ontological middle ground — sharing some things in common with the basic or fundamental (robust causal powers, simple structure, etc.) and some things in common with more complex entities (requiring a certain level of complexity in the structures which cause them, super- vening on other complex structures, etc.). On the standard picture, emergent entities are not the basic building blocks of matter, nor are they complex entities made up of those basic entities — they are some- thing distinct, something which shares features with both the basic and the complex, but cannot really be classed as either. This idea is captured by postulation of fundamental dependent entities. If the paradigm of fundamentality is the fundamental independent entity, then emergent entities have something in common with that para- digm — namely, fundamentality. But they also have something in common with the paradigm of the derivative (the derivative depen- dent entities) — namely, dependence. Fundamental dependent entities carve out much the same ontological middle ground that emergence was originally meant to characterize on the levels picture. So I think it is reasonable to class all fundamental dependent entities as emergent. The issue is at most a terminological one: whether it makes sense to call all and only the fundamental dependent entities ‘emergent’. Since the fundamental dependent entities seem to play the kind of role that we often associate with emergence (and ‘emergence’), I think it makes good sense to do so. Regardless, I think the idea of fundamental dependent entities usefully and non-mysteriously captures the basic ideas that talk of ‘emergence’ is often employed to cover, and that is what seems philosophically important.

4. Examples

Some examples of the sort of ontology that might best be character- ized as emergent will help to make the above picture clearer. None of the examples are in any way meant to be arguments that there is emergence. They are conditional claims (if the world was like this, then there would be emergence) and I am not endorsing the antece- dents. But the conditionals themselves are useful for understanding the characterization given in (OE). I hope to show that (OE) captures a basic idea of emergence, as demonstrated by standard cases. But I also want to argue that the

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 887

characterization of emergence given by (OE) reaches beyond the usual examples of emergence discussed in the literature. You could worry that this means (OE) over-generalizes, but I actually think it is one of its main benefits. Once we see that the same basic phenomenon invoked when talking about consciousness could come up in discus- sions of gunky ontologies or trope bundle theories the phenomenon begins to look less mysterious and more theoretically useful.

4.1 Minds

Take perhaps the most familiar case from the literature on emergence:

the relationship between mind and body. For various reasons philo- sophers are sometimes attracted to the idea that mind and body are distinct. 21 Claiming that mental properties emerge from the complex interaction of physical properties has been one attempt at capturing this distinctness. We can interpret that according to (OE) as follows. Mental properties are part of the fundamental make-up of reality. They do some sort of work (causation, truthmaking, etc.) that cannot be done by anything else. If God wants to create a world like ours, with creatures like us, she cannot just create microphysical particles in certain arrangements and get the mental properties for free. She has to create the mental properties as well. But the existence of mental properties is both caused and sustained by the collective activity of certain physical properties. Without those physical properties persist- ing in very specific arrangements, the mental properties would cease to exist. That is, according to the emergentist, the world of pure Cartesian minds is impossible. Mental properties are dependent:

they depend on the existence of physical properties and relations. Mental properties are thus fundamental dependent entities.

4.2 Living beings and persons

Another familiar case of emergence that can illustrate (OE) comes from the metaphysical theories presented by Trenton Merricks in Objects and Persons and by Peter van Inwagen in Material Beings . Van Inwagen argues that only mereological simples and composites of those simples which are subjects of a life exist; Merricks makes the stronger claim that only simples and composites of those simples which are persons exist. In both cases, compositional nihilism,

21 Often the reasons are causal, though causal reasons do not exhaust the potential motiva- tion. Differences — in persistence conditions, metaphysical structure (you might think minds are ‘simple’ or ‘unified’ in a way bodies are not), etc. — can all be put forward as reasons for a robust mind–body distinction.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

888 Elizabeth Barnes

though thought to be prima facie methodologically attractive, is found wanting. Simples (arranged object-wise) can do all the work we need for ordinary macroscopic objects. But they cannot do all the work, the thought goes, for living beings (van Inwagen) or more specifi- cally for persons (Merricks). These entities have special properties, causal powers, etc., that simply cannot be accounted for solely with reference to simples. But they certainly are not ontologically inde- pendent. They depend for their existence on the fact that some simples are arranged organism-wise; for example, at each moment of a person’s existence there must be some simples arranged qua human organism to sustain the existence of that person. Thus, it seems that on a van Inwagen/Merricks-style ontology, living beings and/or persons are fundamental, even though they are not independent.

4.3 Gunk Examples of (OE) do not come only from familiar discussions of emergence. A gunky ontology might also give substantial motivation for endorsing (OE), simply because a gunky ontology would osten- sibly be one in which nothing is ontologically independent. Assuming that complex objects depend on their parts, if everything has proper parts everything will be dependent. And a gunky ontol- ogy is one in which there is no smallest, indivisible component — everything has proper parts. Thus if the believer in gunk is going to say that anything in her ontology is fundamental, she will need to say that there are fundamental dependent entities. 22 In a gunky ontology, all fundamental entities are fundamental dependent entities. There would be multiple ways to characterize such entities in a gunky framework. A defender of gunk could maintain, for example, that a certain class of objects and everything which composes those objects counts as fundamental. For example, she could claim that electrons and all the infinitely many proper parts that compose elec- trons are fundamental. She still needs to accept fundamental depen- dent entities to say this, because any electron or electron-part she picks

22 Or be a monist. Notice how the picture of emergence given in (OE) gives a nice way of resisting the argument from gunk to monism given in Schaffer 2010 . Schaffer motivates priority monism as the only way of capturing fundamentality within a gunky framework. But if we have reason to accept fundamental entities which are dependent, then there is an alternative way of allowing for fundamentality in a gunky ontology.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 889

out will count as fundamental, but will also be dependent on the parts which comprise it. Alternatively, the gunk theorist could maintain that only a certain class of objects — electrons, living beings, ordinary objects, what- ever — are fundamental. These fundamental entities could combine to compose further derivative entities, but they are also themselves composed of derivative entities. Perhaps, for example, living beings are fundamental but their non-living parts are not. There are derivative complex entities which have living beings as parts — a sum of two living beings, for example (or an ecosystem, for a less rarefied exam- ple). But living beings are also composed of non-living parts — infi- nitely many, if the world is gunky. This view maintains that living beings are fundamental but their non-living parts are derivative: and yet living beings are dependent on their non-living parts. The reason, on this view, that my vital organs are in the world is because I am in the world — my existence is ontologically prior to the existence of my non-living parts. And yet I am dependent on my non-living parts: an entity like me has to have parts like that in order to exist. There is nothing contradictory in this thought (that a fundamental entity could be composed of derivative entities) since the fundamentalist ontology does not commit to fundamentality residing only in the very smallest things. This is particularly relevant to a gunky ontology, in which ‘very smallest things’ is just a case of reference failure. On this picture, there are fundamental entities which are not only dependent (on their parts), but dependent on derivative entities (since their parts are deri- vative). This might strike many as problematic, but it is plausible that when dealing with gunk we often come across things which appear problematic not because of any internal tension, but because we are so accustomed to thinking about these issues within an atomistic framework. To stretch this position yet further, it should be noted that dependence, as characterized, allows that, for example, both an object and its parts could be dependent: that is, nothing we have said so far commits us to dependence being asymmetric. So nothing (at least nothing said here) rules out that a fundamental gunky object and its parts could be dependent. It may be that if these kinds of dependence are allowed, then ‘dependence’ is no longer an appropriate name — maybe something like ‘co-dependence’ or ‘ontological symbiosis’ is more apt — but that question is at most terminological. The point here is simply that this option is not obviously ruled out.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

890 Elizabeth Barnes

4.4 Trope theory

Examples of fundamental dependents can also come from trope theory. Let us assume, as many commonly do, that certain properties cannot exist independently — for example, that there cannot be ‘bare mass’. 23 Mass tropes are arguably dependent — certain other tropes (shape and size tropes, perhaps) must exist, and relate to each other in specific ways, in order for a mass trope to exist. But this should not undermine the idea that a mass trope is a fundamental existent. Mass is something quite distinct from shape and size, and ostensibly cannot be fixed just by shape and size. It is just that it cannot exist indepen- dently of them. The problem of ‘bare mass’ would thus, I contend, be a case of dependent fundamentality. If God is creating the world, she cannot create mass just by creating shape and size (mass does not come for free). But the existence of mass tropes is dependent — you cannot have mass tropes without shape and size tropes. And again, this might plausibly be a case of multiple dependence: bare size or bare shape is likely to be just as bad as bare mass. Alternatively, a trope theorist might maintain that, for example, the shape-size-mass com-

plex (the trope ‘bundle’) is also fundamental. It does work as a unit that the individual tropes cannot do. In that case, she could maintain that each of the mass, shape, and size tropes, though they are them- selves fundamental, cannot exist without the shape-size-mass bundle, which is itself fundamental. Regardless of the specifics, the basic point is this: if you are a trope theorist, properties like mass can thus be understood, via (OE), as emergent.

4.5 Physics

Finally, there are the cases — increasingly appealed to — from quan- tum mechanics wherein the properties of a system cannot, apparently, be exhaustively explained by properties of the parts of that system. 24 That is, the system as a whole seems to display properties which can only be accounted for by the whole, not by the behaviour of the parts. The interpretations of cases like these will obviously be controversial, and I do not want to be committal about them. But they are worth thinking about, even if only to decide what would be the correct thing to say if cases like these were part of a correct physics.

23 Though see Schaffer 2003b for critical discussion.

24 See, for example, Ch. 2 of Maudlin 2007 for discussion.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 891

Purported examples include quantum entanglement and the gen- eration of a field from constituent particles (or the generation of the particles from the field, depending on which interpretation of fields you adopt — likewise for the subsequent examples). In cases such as these the system as a whole seems to be dependent on the relevant components; you only get the holistic system because you have the parts interacting in a certain way, and the parts have to keep inter- acting in a very specific way in order for the system to continue to exist. For example, you only get the field because you have (and continue to have) particles. But once you get the system, you get something, as it were, ‘over and above’ the parts. Again, in the field case, the field is generated by the collective activity of the particles, but it has causal powers that are more than just the collective causal powers of the particles. The system does causal work, has explanatory power, etc., holistically. Thus you are going to need the system itself, rather than just its components, as a truthmaker; likewise, only the system, not the collective actions of its parts, makes it true that the system exists. 25 So the system, on the characterization offered here, is part of fundamental ontology, even though it is ontologically depen- dent on its components. And this seems intuitively correct. If a system, x , is as a whole really something over and above its parts, then ostensibly it would not be enough, if God wants to create system x , for her to say ‘let there be the parts arranged x-wise’. She needs to create the parts and the system. But the system, even though it is part of the fundamental make-up of the world, is still dependent on there being parts that compose it. And so, on the characterization given here, such systems would be apt examples of emergence. 26

5. Advantages

Now that I have given some potential examples of (OE), let me explain why understanding emergence this way is helpful. In what follows,

25 It might be objected that parts + laws are the only things needed for truthmakers in this case, so the system is not a candidate for fundamentality. I am, however, assuming that laws are not components of ontology that can be used as parts of truthmakers (i.e. laws are not themselves fundamental). Moreover, even if a more robust account of laws is assumed, it is far from clear that in such quantum cases (and for instances of emergence in general), you would be able to get anything like the sort of deterministic law you would need for truthmaking.

26 This gives another way of resisting one of the arguments for monism (the argument from physics) from Schaffer 2010.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

892 Elizabeth Barnes

I will argue that the major objections traditionally directed against emergence do not apply, or do not apply with nearly as much force,

to emergence as characterized by (OE). 27 Many common objections to

emergence seem to be objections to emergence couched in a very specific framework (an ontology of ontological ‘levels’); when you place emergence in a fundamentalist framework, those objections no longer look forceful. But first, some caveats. These are simply some of the most common objections put toward emergence — that I discuss my account’s way of getting out of them does not mean I think they have force against all other accounts of emergence, nor does it mean that I think there might not be other equally effective ways of avoiding them. Additionally, these are objections that get tossed out at views labelled ‘emergence’. But it would be a mistake to think that all philosophers developing theories of ‘emergence’ are trying to talk about exactly the same thing — as discussed above, the usage of ‘emergence’ in philo- sophical literature is both ambiguous and sometimes unclear. Consequently, in so far as these are objections, they are only objections to some views which get labelled ‘emergence’. The point here is just that these perennial worries about (some forms of ) emergence — worries which are bound to come up when you start talking about emergence — do not apply to emergence as characterized by (OE).

5.1 Levels ontology

I have described emergence using a ‘fundamentalist’ framework.

Emergence is not usually described with reference (explicit or other- wise) to a meta-ontological framework, but we can find in the back- ground of some discussions of emergence, both positive and critical, an alternative conception of fundamentality. Emergence is sometimes discussed within the context of a picture that divides reality into a series of stratified ‘levels’. This ‘layered’ conception of reality played a key role in the original development of emergence as a philosophical thesis 28 and has been characterized as ‘[providing] an essential

27 Defenders of emergence, of course, have ways of responding to these objections without invoking anything like (OE) (see, inter alia , Wilson 2005 and Crane 2001 for discussion of potential responses). Nothing in the discussion that follows assumes these responses fail. The point is simply that the characterization of (OE) gives a very neat and unified way of dissol- ving a lot of the standard objections to emergence.

28 See especially McLaughlin 1992 for an excellent historical overview.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 893

framework needed to formulate the emergentist/reductionist debate’ (Kim 1999 , p. 19 ). Unsurprisingly, discussions of emergence often uti- lize it heavily. 29 What does the levels picture involve? Again, just like ‘emergence’, its usage is varied and ambiguous, but one way to understand it, a way that seems quite common in the literature and which makes it expli- citly relevant to meta-ontology, is as directly connected to fundamen- tality. It is this specific view — stratified ontological levels, plus facts about how those levels determine fundamentality — that I am label- ling the ‘levels ontology ’. 30 Jaegwon Kim, one of its clearest proponents, explains the basic idea:

The natural world is stratified into levels, from lower to higher, from the basic to the constructed and evolved, from the simple to the more complex. All objects and phenomena each have a unique place in this ordered hierarchy. (Kim 1999 , p. 17)

At the bottom or ‘fundamental’ level are the basic components of matter. It is generally agreed that it is up to the physicists to tell us what these are, on the assumption that the physicists will (or a per- fected physics would, even if actual physics cannot) eventually come up with some particles that are indivisible — the ultimate constituents of the universe. This lowest level is where we find ‘absolute fundamentality ’. 31 The levels then develop in increasing complexity, with each higher- order level made up of those things from the level directly below. Kim again explains:

It is generally thought that there is a bottom level, consisting of whatever microphysics is going to tell us are the most basic physical particles out of

29 See, inter alia , Chalmers 2006, O’Connor and Wong 2005 , Kim 1999 , McLaughlin 1992 , and Wilson 2005 .

30 I am not claiming that all talk of ontological levels commits one to a ‘levels ontology ’, nor am I claiming that it is always assumed in discussions of emergence, nor that it is the sole source of problems for emergence. Rather, I am claiming that it is one common background (either explicit or, perhaps more often, implicit) in discussions of emergence, one which is usually contrasted with the fundamentalist background I have described above. This is the notion of ontological levels that is, for example, criticized in Schaffer 2003 and defended in Callender 2001.

31 The levels picture is sometimes explained in terms of fundamentality, sometimes in terms of basicness or simplicity, and sometimes with some combination of these terms. I am assum- ing that something like a notion of fundamentality is in play in most discussions of the levels pictures (see Schaffer 2003 ).

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

894 Elizabeth Barnes

which all matter is composed … [At] higher levels we find structures that are made up of entities belonging to the lower levels. (Kim 1993 , p. 337 )

So, for example, if subatomic particles are at level n, then atoms will be at n + 1 , molecules at n + 2 , and so on. And this levels structure is then explicitly connected to fundamentality. Things get more complex, and as a result less fundamental, the further up the levels hierarchy you go. Atoms exist and tables exist, but atoms are more fundamental than tables. And neither is absolutely fundamental. This talk of levels is, of course, a metaphor, and it must not be taken too seriously. The important thing to remember, and what the strati- fied levels picture is meant to illustrate, is that there are degrees of fundamentality which vary inversely with degrees of complexity. So it is not just that the world is stratified into ontological levels — it is that this levels structure (perhaps necessarily) determines the facts about fundamentality. The absolutely fundamental is the simple. As things get more complex, they get less fundamental. 32 To recap: the basic tenets of what I am calling the ‘levels ontology ’ are that (i) there is a stratified hierarchy of ontological levels; (ii) this hierarchy determines the facts about fundamentality. As discussed in section 1, this ontological framework contrasts markedly to what I have called the fundamentalist ontology. The fundamentalist ontology can certainly allow for a world in which all and only the smallest things are fundamental, with those things combining to compose other, derivative things (that is, the fundamentalist ontology can recapture a structure that looks exactly like the structure of the levels ontology). But it does not require this sort of picture. And it does not allow for degrees of fundamentality. In what follows, I will argue that many of the problems traditionally associated with emer- gence are problems for emergence couched in a levels ontology, and that they dissipate when that framework is replaced with the one advocated here.

5.2 Causation Emergent causation is problematic. Emergent entities are usually understood as having unique, non-redundant causal powers which can affect not only other emergent entities, but also the entities (or kinds of entities) from which they emerge (‘downward causation’). This is often considered problematic, especially by those who endorse a broadly physicalist account of causation. Emergent entities are not,

32 For detailed discussion of these issues, see Schaffer 2003 .

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 895

ex hypothesi , part of the very basic building blocks of matter (they emerge, further down the line), yet it is often assumed that all causa- tion can be explained solely in terms of those basic building blocks. So if emergent entities have causal powers — and not only that, but causal powers that affect what the basic physical things do — then we have a problem. 33 Notice, however, that this problem is couched in the very specific ontological framework of ontological levels. And that framework is explicitly rejected here. The claim — meant to be in tension with emergence — is that all causation must ultimately be accounted for by what is absolutely fundamental. On the levels picture, those are the ultimate constituents of matter (the very smallest things). Since emer- gent entities explicitly are not those kinds of entities, we run into problems saying that emergent entities have real, non-redundant causal powers — particularly causal powers that can impact the activ- ity of the fundamental bits of matter. Contrast this to the background ontological characterization that informs (OE). Fundamentality, on the levels picture, is something which can come in degrees, and which will always be found to the highest degree in the smallest, most basic things there are. But on the characterization of fundamentality given here, fundamentality does not come in degrees (things are either fundamental or they are not, in which case they are derivative) and we have no guarantee that the basic, smallest things will be what is fundamental (e.g. the monist could be right, in which case the very biggest thing is fundamental, or the gunk theorist could be right, in which case there is no class of things which are the very smallest). Emergent causation is a problem if you situate it within a levels ontology. Once we have levels in place, the physicalist then claims warrant to explain all causation solely in terms of what is absolutely fundamental (the very basic things), which will never include emer- gent entities. But the analogous claim for the ontological structure assumed here looks to be this: all causation can be explained solely in terms of what is fundamental. 34 In that case, of course, there is no

33 See especially Kim 1999 — note that this is simply an emergence-specific formulation of his famous ‘exclusion argument’ (as in, inter alia , Kim 1993 ).

34 Since there is no notion

of

‘absolute’ — as contrasted with partial — fundamentality.

Things are either fundamental or they are not, in which case they are derivative.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

896 Elizabeth Barnes

causation problem for emergence, since emergent entities are funda- mental (just not independent). You might think that the analogous claim should instead be: all causation can be explained solely with reference to independent enti- ties. But this seems to be too strong. Surely we think that there could still be causation in a gunky world, or that a mass trope can have causal powers, etc. Restricting to independent entities in this way looks unmotivated. Some philosophers further maintain that all causation should be explained only with reference to physical entities. But this is not a problem for (OE) either. There is no pressure, on the theory of emer- gence offered here, to think that emergent entities are non-physical. The account is neutral in this respect. There is some thought, again hostage to a ‘levels’ ontology, that emergence would have to involve non-physical ontology, because all the physical ontology is exhaus- tively explained by the very basic things (the absolutely fundamental), and the emergent entity represents ‘something new’ over and above these. But again, the analogous claim on the characterization given here simply looks to be that all physical ontology is exhaustively explained by fundamental ontology, which is perfectly compatible with some of it being explained by dependent fundamental (i.e. emer- gent) ontology. 35

5.3 Ontological ‘weight’ Another standard problem for emergence comes in saying why, exactly, we need to treat emergent entities with ontological serious- ness. Emergent entities, on the levels picture, are not absolutely fun- damental — they are not the basic, indivisible building blocks of matter 36 — or at the very least they are very different from the other sorts of things that are absolutely fundamental. But they are also not just another variety of complex entity, because they are somehow

35 Indeed, taking recent science to heart should at least suggest that it is not outlandish to suppose that not all of the fundamental physical entities are ontologically independent. The atomistic picture of tiny little independent bits of matter bumping into one another is elegant, but we do not have any a priori warrant that it is accurate.

36 Defenders of emergence might protest that emergent entities are ‘basic’. But this seems a terminological shift. If we take ‘basic’ to mean something like ‘the smallest/indivisible/out of which all other things are built’, then emergent entities do not count as basic. If basic is taken as a primitive which can come apart from mereological simplicity — something more like fundamentality — then we have moved away from the levels ontology. Once basicness or fundamentality comes apart from mereological structure, then, as Schaffer comments, ‘the entire “levels” metaphor is perhaps best abandoned’ (Schaffer 2003 , p. 500).

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 897

meant to carry more ontological weight. 37 So the question becomes:

Where do we situate them within the hierarchy of levels? Are they a higher-level entity that somehow has unique causal powers not usually possessed by higher-level entities? Are they higher-level entities picked out by their unique structure (e.g. they are simple, whereas most higher-level entities are complex)? Are they perhaps completely sui generis — not fitting into the stratified layers of the physical world because they are non-physical? It is difficult to say what exactly emer- gent entities would be, and where in the layers of reality we should place them. The problem dissipates once we divorce fundamentality from the atomic structure of the levels picture. There is no question, on the account characterized by (OE), of why we take emergent entities to have ontological ‘weight’ — they are fundamental. There is, of course, the question of why we should take them as fundamental. But that is

a question for independent theorizing: what kind of theoretical

work they do, whether we need them as truthmakers, etc. So we would need to think hard about the entities themselves — whether they are what God has to include in her ontological shopping basket

or whether they must be included on a basic list of truthmakers, etc.

But these are just the basic sorts of questions asked of any entity (dependent or otherwise) in deciding whether or not we should con- sider it fundamental. There is no special problem about emergent entities per se. In contrast to the levels picture, the notion of fundamentality defended here makes no assumption about what sorts of things are fundamental. That is a task for further investigation, and it would be up to the emergentist to argue that some ontologically dependent entities are fundamental. The point here is simply that treating such (dependent) entities with ontological seriousness is not automatically problematic, because the notion of fundamentality in play does not constrain her into considering only certain kinds of entities as candi- dates for the fundamental.

5.4 Non-reductive physicalism Discussions of emergence often struggle with the question of whether and to what extent emergence is distinct from non-reductive

37 Alternatively, sometimes emergentism is described as each level being emergent from the previous one (as in, e.g., Kim 2006 ). It is not clear that this formulation would make the problem any better.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

898 Elizabeth Barnes

formulations of physicalism. 38 This debate is murky for many rea- sons — most notably because there is wide variance in what is meant by ‘emergence’ and wide variance in what is meant by ‘non- reductive physicalism’. 39 However, if we understand ‘emergence’ via (OE), it is easy to see how emergence (so characterized) is distinct from at least some standard construals of non-reductive physicalism. 40 Non-reductive physicalism is often understood as a claim about explanation. Consider the layered hierarchy of the levels ontology. Now consider facts of type F at level n and (corresponding but more basic) facts of type G at level n 1 . Are there so-called bridge laws of the form ‘F iff G’? If there are such bridge laws, then we can reduce facts about things at level n (F facts) to facts about things at level n 1 (G facts). That is, we can explain everything there is to explain about the Fs by talking about the Gs. So if, for example, facts about chemistry are reducible to facts about physics, we could explain (with a completed physics) everything about chemistry using only facts about physics; we would not need recourse to an additional, sui gen- eris set of chemistry facts. If we cannot provide such bridge laws, then we cannot provide the relevant explanations and our theory is non- reductive (i.e. you cannot reduce facts at level n to facts at level n 1 , for at least some levels n and n 1 ). Physicalism — understood according to the levels ontology — is (roughly) the doctrine that all facts about the world are fixed by the facts at the most fundamental level. Thus one simple way of under- standing non-reductive physicalism is this: it combines the ontological thesis of physicalism with the explanatory thesis of non-reduction. All facts are fixed by the facts at the most fundamental level, but not all facts are explained by the facts at the most fundamental level. The relevantly similar claim, in a fundamentalist ontology, would be two-fold. Firstly, that all fundamental facts are physical facts. Secondly, that all facts simpliciter are fixed by the fundamental facts, even though not all facts are explained by the fundamental facts.

38 Some (e.g. Crane 2001 ) take them to be the same. Some (e.g. Chalmers 2006 ) take there to be stronger and weaker forms of emergence, with the weaker forms of emergence identical to non-reductive supervenience. Some (e.g. Kim 2006) argue that they are distinct, but suffer from similar problems. And some (e.g. Wilson 2005 , O’Connor and Wong 2005 ) construe them as very different.

39 See Wilson MS for an excellent and exhaustive taxonomy of the issues in play in this debate.

40 And perhaps others as well, though I do not have the space here to go into a detailed discussion of interpretations of non-reductive physicalism.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 899

The fundamentalist can easily accept this — it is no part of her theory that anything like bridge laws can be provided to furnish properly informative explanations of facts about derivative entities in terms of facts about fundamental entities. Just because, for example, the truthmaker for ‘Tables exist’ is the existence of simples rather than the existence of tables does not entail that you can explain everything about tables just by talking about simples. Ontological fundamental- ism is, by itself, completely neutral with regards to whether reduction- ism is true. The important point here, though, is that the claim that not all facts are explained by the fundamental, even though all facts are fixed by the fundamental, is completely separate from the claim that there is emergence. Commitment to emergence, as understood via (OE), is simply commitment to fundamental dependent entities. Whether there are any such entities seems orthogonal to the question of whether fundamental facts exhaustively explain derivative facts. So if emergence is construed according to (OE), the question of whether and how it is distinct from at least one standard interpretation of non- reductive physicalism becomes a non-starter. Another interpretation of non-reductive physicalism is simply as the denial of type physicalism — that is, the denial that, for any prop- erty F, that property is identical to some physical property G. We can still have token physicalism — that any instantiation of a property is identical to the instantiation of some physical property (where that claim is de dicto , not de re ) — but the possibility of multiple realiz- ibility precludes reduction. 41 Emergence, on the account presented here, is consistent with this type of non-reductive physicalism, but it does not entail it and is not entailed by it. And that is simply because this account of emergence is completely neutral as to whether emergent entities are non-physical. You could have a world in which there are fundamental dependent entities, but in which type physicalism is true (i.e. in which all proper- ties are physical properties). Again, on this account the question of whether non-reductive physicalism is true looks to be completely orthogonal to the question of whether there is emergence.

41 An instantiation of property F, for example, might be identical to the instantiation of physical property G at t , but an instantiation of F at t * be identical to physical property H. So we cannot say that F = G nor that F = H.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

900 Elizabeth Barnes

6. Conclusion

I have argued that within a fundamentalist ontology (which makes a core distinction between fundamental and derivative entities) there is room for a further distinction: that of dependent and independent entities. I maintain that this set of distinctions can usefully explain the mysterious idea of ontological emergence. The characterization of emergent entities as simply those entities which are fundamental but not independent is simple (much simpler than many of the usual explanations) and avoids many of the standard problems associated with emergence. There may well be other meta-ontological frame- works that can provide similarly useful explications of emergence — whether that is the case is beyond the scope of this paper. The point here is simply that carefully considering the ontological background in which emergence is situated can help clarify it, and that a fundamen- talist ontology yields a natural and attractive account of emergence. 42

References

Beckerman, Ansgar 1992 : Emergence or Reduction? Gruyter: New York. Bueno, Otavio 2009: New Waves in Philosophy of Mathematics . New York: Palgrave Macmillan. Callender, Craig 2001: ‘Why Be a Fundamentalist?: Reply to Schaffer’. Pacific APA, San Francisco, March 2001 < http://philpapers,org/>. Cameron, Ross 2008: ‘Truthmakers and Ontological Commitment’. Philosophical Studies, 140 , pp. 1 18 . —— 2010: ‘Quantification, Naturalness, and Ontology ’, In Hazlett 2010, pp. 826. —— 2010b: ‘Necessity and Triviality ’. The Australasian Journal of Philosophy , 88 , pp. 40115 . Clayton, Philip 2006 : The Re-Emergence of Emergence . Oxford: OUP. Crane, Tim 2001: ‘The Significance of Emergence’, In Gillet and Loewer 2001, pp. 207 24 . Chalmers, David 2006 : ‘Strong and Weak Emergence’, In Clayton 2006, pp. 244 55 . Fine, Kit 2001 : ‘The Question of Realism’. Philosophers’ Imprint , 1 , pp. 130 . Gillett, Carl 2001 : Physicalism and Its Discontents . Cambridge: CUP.

42 Many thanks to Ross Cameron, Laurie Paul, Jonathan Schaffer, Jason Turner, Robbie Williams, and Jessica Wilson. Thanks also to two anonymous referees.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013

from at Rutgers University on January 19, 2014http://mind.oxfordjournals.org/Downloaded

Emergence and Fundamentality 901

Hazlett, Alan 2010: New Waves in Metaphysics . New York: Palgrave Macmillan. Heil, John 2003: From an Ontological Point of View. Oxford: OUP. —— 1993 : Mental Causation . Oxford: Clarendon Press. Kim, Jaegwon 1999 : ‘Making Sense of Emergence’. Philosophical Studies, 95 , pp. 336 . —— 1993 : ‘The Non-Reductivist’s Trouble with Mental Causation’, In Heil 1993 , pp. 189210 . —— 1992 : ‘Downward Causation in Emergence and Non-Reductive Physicalism’, In Beckerman 1992 , pp. 11938 . Langton, Rae and David Lewis 1998 : ‘Defining “Intrinsic”’. Philosophy and Phenomenological Research , 58 , pp. 333 45 . Marras, Ausonio 2006: ‘Emergence and Reduction: Reply to Kim’. Synthese , 151, pp. 5619 . Maudlin, Tim 2007: The Metaphysics Within Physics . Oxford: OUP. McLaughlin, Brian 1992 : ‘The Rise and Fall of British Emergentism’, In Beckerman 1992 , pp. 49 93 . Merricks, Trenton 2003: Objects and Persons . Oxford: OUP. O’Connor, Tim and Hong Yu Wong 2005: ‘The Metaphysics of Emergence’. Nous, 39 , pp. 65878 . Rayo, Agustı´n 2009 : ‘Toward a Trivialist Account of Mathematics’, In Bueno 2009, pp. 239 63 . Schaffer, Jonathan 2003 : ‘Is There a Fundamental Level?’ Philosophical Studies, 37 , pp. 498 517 . —— 2003b: ‘The Problem of Free Mass: Must Properties Cluster?’ Philosophy and Phenomenological Research , 66 , pp. 125 38 . —— 2010: ‘Monism: The Priority of the Whole’. The Philosophical Review, 119, pp. 31 76. —— 2010b: ‘The Internal Relatedness of All Things’. Mind , 119, pp. 34176. Van Inwagen, Peter 1990 : Material Beings . Ithaca, NY: Cornell University Press. Williams, J. R. G. 2010 : ‘Fundamental and Derivative Truths’. Mind , 119, pp. 10341 . Wilson, Jessica 2005 : ‘Supervenience-Based Formulations of Physicalism’. Nous, 29 , pp. 426 59 . —— MS: ‘Metaphysical Emergence: Weak and Strong’.

Mind, Vol. 121 . 484 . October 2012

Barnes 2013